Open Source Electric Supercharger

[Tenthousanddays]

Originally Posted by neutron256

I'm not that interested in the business side, but I love the idea of an off the shelf motor/compressor that can be adapted by tuners to various platforms and I'm more then willing to help be the proof of concept. I've already got a pretty solid charging/control system, it's just my motor and compressor that are troublesome so I'd love to test out this design.

Awesome. I'll get the details for cost/availability of one of the direct drive systems for you to play around with. Hopefully I can get you one at less than retail pricing.

In the meantime I'll get back in touch with the folks over at TurboKits and gauge their interest in adding this to their already extensive line up of kits.

[WNDSRFR]

PHANTOM

[Tenthousanddays]

Some answers for you Neutron:

Is the motor controller integrated with the assembly? Our low voltage brushless motors, including the one upon which the direct drive supercharger is based, have integrated commutation electronics that allow them to operate directly from a battery, in this case 24V. However, speed control requires the application of variable DC voltage to the input terminals, which means a buck regulator or DC motor controller with an external inductor for filtering high frequency PWM will be needed.

What is the cost? Less than $1,500 for the direct drive model, whereas the much higher speed and pressure version with an integrated gearbox will be well over $2,000.

How soon will it be available for purchase? Within six to eight weeks, possibly sooner, if early adopters are willing to make commitments and push for it.

To my knowledge, no special machining will be needed, in order to use our centrifigul superchargers.

Dan

[Tenthousanddays]

I'm optimistic! No specs yet - in terms of size, power consumption, cfm ETC... It'd be awesome to mount this up with what you're already using (unichip/ESC/charging device/batteries) and get the 7-9 PSI he say's it'll put out.

I'd definitely like to join in on setting this up, but I'd like to wait until the specs have been posted. I hope it's not much bigger than what you're using, as then we're looking at compressor relocation. Being able to utilize the Unichip setup would be optimal as well.

Was looking through the thread again, and had a question - what made you go for the Odyssey batteries as opposed to the Power-Sonic ones from the Phantom system?

[neutron256]

Originally Posted by Tenthousanddays

I'm optimistic! No specs yet - in terms of size, power consumption, cfm ETC... It'd be awesome to mount this up with what you're already using (unichip/ESC/charging device/batteries) and get the 7-9 PSI he say's it'll put out.

I'd definitely like to join in on setting this up, but I'd like to wait until the specs have been posted. I hope it's not much bigger than what you're using, as then we're looking at compressor relocation. Being able to utilize the Unichip setup would be optimal as well.

Was looking through the thread again, and had a question - what made you go for the Odyssey batteries as opposed to the Power-Sonic ones from the Phantom system?

I think people go with the Power-Sonics for the Phantom setup because they are more affordable. I think at least one person went the the Odysseys like I did. odyssey is just much better quality, and I really like that it has screw terminal connectors rather then tabs.

[AN]

Ad one more to the list of members who joined in with the sole purpose of following this thread ... I am an MX-5 owner and really hope to see a DIY kit come out of this effort.

Great job neutron256 ... I hope you get back to it soon ... I had a couple of ideas / questions that I wanted to share and see if you have done some thinking / research about it ...

> For power delivery ... how about using a battery pack in conjunction with (parallel) some large caps ... This setup should be able to provide high peak current demands of any hi-power motor ..

> also, for tuning ... while closed loop would be ideal ... but given how well the Phantom solution is working with and on-off switch .. I think the three mode solutions is rather simple and elegant ...

1. TPS less than 25% off
2. TPS between 25% ~ 75% ... Ramp up the motor with the rpm
3. TPS 75% + run full speed
4. Also if the motor cut-off should be slightly before rpm redline ... to avoid surge (only in picture if we have a 7+ psi eSC.

Apologies if this these have been covered before ...
Also .. really hoping the LM5 motor/compressor setup is available soon and priced sub $1000... sounds like an ideal solution for this setup ...

Does anyone know what are the specs on the motors used in the Phantom setup ...
Thanks all.

[neutron256]

Originally Posted by AN

Ad one more to the list of members who joined in with the sole purpose of following this thread ... I am an MX-5 owner and really hope to see a DIY kit come out of this effort.

Great job neutron256 ... I hope you get back to it soon ... I had a couple of ideas / questions that I wanted to share and see if you have done some thinking / research about it ...

> For power delivery ... how about using a battery pack in conjunction with (parallel) some large caps ... This setup should be able to provide high peak current demands of any hi-power motor ..

> also, for tuning ... while closed loop would be ideal ... but given how well the Phantom solution is working with and on-off switch .. I think the three mode solutions is rather simple and elegant ...

1. TPS less than 25% off
2. TPS between 25% ~ 75% ... Ramp up the motor with the rpm
3. TPS 75% + run full speed
4. Also if the motor cut-off should be slightly before rpm redline ... to avoid surge (only in picture if we have a 7+ psi eSC.

Apologies if this these have been covered before ...
Also .. really hoping the LM5 motor/compressor setup is available soon and priced sub $1000... sounds like an ideal solution for this setup ...

Does anyone know what are the specs on the motors used in the Phantom setup ...
Thanks all.

Thanks for the support.

I have a set of UltraCaps that I got specifically for this project and will probably try to run with the batteries at some point. First I need to get the basic setup a lot more solid.

Personally I'm not so concerned about surge at redline. The PSI falls as RPM rises so at redline PSI isn't going to be that high unless someone can build a motor/compressor that can deliver much higher pressures then Phantom is currently seeing.

Currently since I'm using the Unichip module I can easily use it to create a boost control map using TPS and RPM. It is however a very cluttered install and l'd like something cleaner. I'm considering custom designing a controller that takes the TPS signal directly from the throttle body, and the somehow reads RPM.

My thought is the simplest way to read RPM would be to just detect the signal from signal going to one of the plugs. I'd need put a little more thought into the best way to do this.

My biggest problem right now though is redesigning my compressor. It's a lot of power in a very small package. I'm thinking more and more the best way to go is some sort of transmission to use a larger slower motor to obtain the compressor speeds needed. What I really would like is an affordable centrifugal supercharger that I could drive with the electric motor.

[AN]

Originally Posted by neutron256

My biggest problem right now though is redesigning my compressor. It's a lot of power in a very small package. I'm thinking more and more the best way to go is some sort of transmission to use a larger slower motor to obtain the compressor speeds needed. What I really would like is an affordable centrifugal supercharger that I could drive with the electric motor.

Yeah, for proof of concept, you should keep things as simple as possible .. also, if you over-complicate your goals ... you might loose interest (I hope not).

Direct drive would be ideal, I hope the LM5 motor from e-cycle comes through at a reasonable price point ... step up gearing at the rpm involved adds a lot of break-points/cost/ mass / maintenance... If this challenge is handled electrically, in the long run it will have a lot of benefits. However, if you ARE leaning towards getting a centrifugal SC for now, I see a few used Vortechs in the $600 ~ $1000 range on eBay. I even see a Rotrex for $900.

I do not have very good understanding of compressor maps ... (I am a EE major, working in IT :| ) .. should I be looking for a compressor which moves a lot of CFM at relatively low rpms and PRs ? Any resource you can point me to ... that will get me reading ?

Also, In your opinion, whats the easiest way to get a usable digital or analog signal for TPS and RPM ? w/o the use of expensive electronics? Is the throttle sensor a 1~5V signal on these / other modern cars ? RPMs would be pulses I imagine.

One more observation I have from Fenton's thread is that after adding additional (smaller 2x 9Ah) batteries to the 2x 18Ah batteries, they are seeing higher boost levels (0.5 psi) through out. This indicates that the ability to flow current at peak demand is definitely a bottleneck here and the Caps should help with that a lot.

[neutron256]

Originally Posted by AN

Yeah, for proof of concept, you should keep things as simple as possible .. also, if you over-complicate your goals ... you might loose interest (I hope not).

Direct drive would be ideal, I hope the LM5 motor from e-cycle comes through at a reasonable price point ... step up gearing at the rpm involved adds a lot of break-points/cost/ mass / maintenance... If this challenge is handled electrically, in the long run it will have a lot of benefits. However, if you ARE leaning towards getting a centrifugal SC for now, I see a few used Vortechs in the $600 ~ $1000 range on eBay. I even see a Rotrex for $900.

I do not have very good understanding of compressor maps ... (I am a EE major, working in IT :| ) .. should I be looking for a compressor which moves a lot of CFM at relatively low rpms and PRs ? Any resource you can point me to ... that will get me reading ?

Also, In your opinion, whats the easiest way to get a usable digital or analog signal for TPS and RPM ? w/o the use of expensive electronics? Is the throttle sensor a 1~5V signal on these / other modern cars ? RPMs would be pulses I imagine.

One more observation I have from Fenton's thread is that after adding additional (smaller 2x 9Ah) batteries to the 2x 18Ah batteries, they are seeing higher boost levels (0.5 psi) through out. This indicates that the ability to flow current at peak demand is definitely a bottleneck here and the Caps should help with that a lot.

I do love the simplicity of a direct drive system, but it has the drawback that we're looking at speeds that really reach the upper limit of what these hobby brushless motors are capable of. One thing I noticed in my testing is the motors heat up quickly when pushed to the limit.

The reason I think fitting an electric motor to an off the shelf centrfegual super charger is that:
1. I'm not going to be able to build a better transmission myself
2. In addition to the step up gearing, the compressor wheels/housings are designed to produce more boost at slower speeds then a similar sized turbo.

The Phantom setup is actually a custom machined turbo housing to fit a larger compressor wheel from what I understand. Check YouTube for videos explaining compressor maps there are several out there.

TPS signals are 0-5V so easy enough to read with any micro controller. My thought for RPM is to detect the pulses going to one of the spark plugs and then have a micro controller read that. The FA20 engine uses a plug on coil setup so I'm not sure what the signaling voltage triggering the coil looks like but I'm sure it could be figured out. Some people had mentioned reading it off the OBD port but the data there has some lag due to the CAN bus refresh rate. Plus That sort of interface design is a bit beyond my programming skills.

I have no doubt that current is a bottle neck. It will be interesting to see how much caps would help. They are best at handling short spikes in current so it's going to take some very large caps to help for more then just quick bursts.

Keep the ideas coming. I love getting other peoples thoughts and brainstorming.

[AN]

Originally Posted by neutron256

I have no doubt that current is a bottle neck. It will be interesting to see how much caps would help. They are best at handling short spikes in current so it's going to take some very large caps to help for more then just quick bursts.

Keep the ideas coming. I love getting other peoples thoughts and brainstorming.

Yeah , I was seeing that huge caps are rather affordable now ... Two of these guys would provide 60 Farad ... that's a lot to power the motor for a few seconds all by themselves. In conjunction with an AGM battery .. it might just be a perfect combo (in theory)

http://www.bestbuy.com/site/power-ac...at165900050030

[neutron256]

Originally Posted by AN

Yeah , I was seeing that huge caps are rather affordable now ... Two of these guys would provide 60 Farad ... that's a lot to power the motor for a few seconds all by themselves. In conjunction with an AGM battery .. it might just be a perfect combo (in theory)

http://www.bestbuy.com/site/power-ac...at165900050030

I agree I think it's got a lot of potential.

I'd worry about adding more stuff to the system. When I had batteries, charger, Unichip, Brushless controller, motor/compressor my engine bay was starting to get very crowded and a mess of wires and connections. It offended my brains sense of tidiness a bit, not to mention the added weight.

A lot of this could be solved by moving the power system to the trunk as some have with the Phantom ESC.

I just personally love clean and simple and if it grows into too much of a beast my practical side would rather just have a conventional supercharger. (Not that it would stop me from trying it just for curiosity sake)

[neutron256]

I did some searching on ebay yesterday for supercharger options. I found an interesting supercharger for SeaDoo 4-tec engines. The gearing is integrated into the engine but the compressor itself is designed with a large wheel designed to turn at slower speeds than compressors from turbos.

I couldn't find any compressor maps but after some digging and some math it looks like on a 1.8L engine they produce about 14psi at 8100 RPM. The gear ratio works out to be about 43k RPM on the compressor wheel.

I wouldn't expect to get anywhere near 14psi, but that lends to much more reasonable motor speeds for a direct drive from the motor. Also the larger wheel gives me more options when it comes to coupling the compressor wheel to the motor shaft.

I went ahead and ordered a used SeaDoo compressor to experiment with. There seem to be a lot of thes on eBay. The bearings/clutch setup frequently needs to be rebuilt probably due to operating in a marine environment. But as long as the wheel hasn't contacted the housing I don't really need the bearing/clutch portion working.

[AN]

Originally Posted by neutron256

I did some searching on ebay yesterday for supercharger options. I found an interesting supercharger for SeaDoo 4-tec engines. The gearing is integrated into the engine but the compressor itself is designed with a large wheel designed to turn at slower speeds than compressors from turbos.

I couldn't find any compressor maps but after some digging and some math it looks like on a 1.8L engine they produce about 14psi at 8100 RPM. The gear ratio works out to be about 43k RPM on the compressor wheel.

I wouldn't expect to get anywhere near 14psi, but that lends to much more reasonable motor speeds for a direct drive from the motor. Also the larger wheel gives me more options when it comes to coupling the compressor wheel to the motor shaft.

I went ahead and ordered a used SeaDoo compressor to experiment with. There seem to be a lot of thes on eBay. The bearings/clutch setup frequently needs to be rebuilt probably due to operating in a marine environment. But as long as the wheel hasn't contacted the housing I don't really need the bearing/clutch portion working.

Great ... can you share a link of the ebay item ... also ... I know you have gone through a few motors but in you opinion .. what is the best motor(s) to couple to this new find ?

I just want to start looking/reading on a possible setup.
Thanks

[rusty959]

Originally Posted by AN

Yeah , I was seeing that huge caps are rather affordable now ... Two of these guys would provide 60 Farad ... that's a lot to power the motor for a few seconds all by themselves. In conjunction with an AGM battery .. it might just be a perfect combo (in theory)

http://www.bestbuy.com/site/power-ac...at165900050030

First of all, 60 farads will lose over 3V each second with the power levels we are looking at.

Furthermore, car audio caps usually have a max voltage of 18-20 volts, so you would only be able to use a single battery (12v) battery with that cap, else put them in series which halves the capacitance. So it would take 4 of those caps to get 30 farads on a 24v setup. 60F would need 8. Therein lies the big issue with running high voltages with caps.